CNShanghai Jiao Tong UniversitySchool of Integrated Circuits (Information and Electronic Engineering)
Weiyao Lin
特聘教授
Works on multimedia analysis, computer vision, multimodal large models, embodied intelligence and low-altitude robots.
Weiyao Lin's lab at SJTU works on embodied & active vision, vision-language & multimodal and multimodal & grounded language. 42 papers since 2023, 26 at top venues, mostly ICLR and AAAI. 34 alumni, most often to Ant Group and Tencent. Applications close .
At a glance
- papers, 2023–2026
- 11/ year
- at T1 venues
- 62%
- PhD students
- 0
- MS students
- 0
- alumni listed
- 34
Papers by yearfrom the lab homepage, 2023–2026
- 15 papers2026 · T1 9 · collecting
- 13 papers2025 · T1 8
- 11 papers2024 · T1 7
- 3 papers2023 · T1 2
T1 venuesOtherVenues: ICLR, AAAI, ECCV, IJCV, NeurIPS
Research focuspaper titles that use each topic's words
- Vision-language & multimodal5 papers
- Multimodal & grounded language5 papers
- Embodied & active vision2 papers
- multimedia analysis0 papers
- multimodal large0 papers
Paper list
From the lab homepage.
202615 papers
- ECCV (T1)Autoregressive Image Generation Needs Only a Few Lines of Cached Tokens
- IEEE Trans. Circuits and Systems for Video TechnologyBridging the Gap between Implicit and Explicit Representation for Efficient Image Compression
- AAAI (T1)CogStream: Context-guided Streaming Video Question Answering
- IEEE Trans. Circuits and Systems for Video TechnologyContending with Depth Ambiguity in Monocular 3D Pose Estimation via Multi-Hypothesis Modeling and SMPL Priors
- ACL (T1)CreditDecoding: Accelerating Parallel Decoding in Diffusion Large Language Models with Trace Credits
- ICLR (T1)DND: Boosting Large Language Models with Dynamic Nested Depth
- IEEE Trans. Circuits and Systems for Video TechnologyFeedforward Compression of Static and Streamable 3D Gaussian Splatting
- IEEE Trans. Pattern Analysis and Machine IntelligenceFew-shot Action Recognition via Intra- and Inter-Video Information Maximization
- AAAI (T1)Head-Aware KV Cache Compression for Efficient Visual Autoregressive Modeling
- IJCV (T1)Looking beyond Visible Cues: Implicit Video Question Answering via Dual-Clue Reasoning
- ICLR (T1)Massive Activations are the Key to Local Detail Synthesis in Diffusion Transformers
- IEEE Trans. Pattern Analysis and Machine IntelligenceMECD+: Unlocking Event-Level Causal Graph Discovery for Video Reasoning
202513 papers
- IJCV (T1)Achieving Procedure-Aware Instructional Video Correlation Learning under Weak Supervision from a Collaborative Perspective
- ICLR (T1)CAKE: Cascading and Adaptive KV Cache Eviction with Layer Preferences
- IEEE Trans. Circuits and Systems for Video TechnologyCSTA: Spatial-Temporal Causal Adaptive Learning for Exemplar-Free Video Class-Incremental Learning
- IEEE Intl. SymEfficient Lossless Compression with Distribution Quantized Finite-State Autoregressive Model
- ICLR (T1)Fast Feedforward 3D Gaussian Splatting Compression
- IEEE Trans. Pattern Analysis and Machine IntelligenceHAC++: Towards 100X Compression of 3D Gaussian Splatting
- Nature Machine Intelligence (T1)Learning vision-based agile flight via differentiable physics
- SCIENCE CHINA Information SciencesLow-rank Winograd Transformation for 3D Convolutional Neural Networks
- ICCV (T1)NeCGS: Neural Compression for 3D Geometry Sets
- ICLR (T1)Rodimus*: Breaking the Accuracy-Efficiency Trade-Off with Efficient Attentions
- RA-LSeeing Through Pixel Motion: Learning Obstacle Avoidance from Optical Flow with One Camera
- IJCV (T1)Toward Accurate and Robust Pedestrian Detection via Variational Inference
202411 papers
- ACM Trans. Multimedia Computing, Communications, and ApplicationsA Unified Framework for Jointly Compressing Visual and Semantic Data
- ECCV (T1)BaSIC: BayesNet Structure Learning for Computational Scalable Neural Image Compression
- AAAI (T1)Collaborative Weakly Supervised Video Correlation Learning for Procedure-Aware Instructional Video Analysis
- Visual IntelligenceControllable Augmentations for Video Representation Learning
- ACM MMDAC: 2D-3D Retrieval with Noisy Labels via Divide-and-Conquer Alignment and Correction
- AAAI (T1)Density Matters: Improved Core-set for Active Domain Adaptive Segmentation
- ICLR (T1)Finite-State Autoregressive Entropy Coding for Efficient Learned Lossless Compression
- ECCV (T1)HAC: Hash-grid Assisted Context for 3D Gaussian Splatting Compression
- NeurIPS (T1)MECD: Unlocking Multi-Event Causal Discovery in Video Reasoning
- ACM Trans. Multimedia Computing, Communications, and ApplicationsScene Graph Lossless Compression with Adaptive Prediction for Objects and Relations
- ECCV (T1)TimeCraft: Navigate Weakly-Supervised Temporal Grounded Video Question Answering via Bi-directional Reasoning
20233 papers
- NeurIPS (T1)BasisFormer: Attention-based Time Series Forecasting with Learnable and Interpretable Basis
- IJCV (T1)HiEve: A Large-Scale Benchmark for Human-Centric Video Analysis in Events
- IEEE Trans. Neural Networks and Learning SystemsSpatio-Temporal Point Process for Multiple Object Tracking
Where alumni went34 listed, 73% to industry
Industry 16Academia 6
Most often to
- Ant Group
- Tencent
- Alibaba
Also to
- Shanghai Jiao Tong University
- Chinese Academy of Science
- East China Electric Power Design Institute
Apply to SJTUFall 2027
Computer Science PhD (international applicants)
- Round
- Self-funded
- Opens
- —
- GRE
- See program page
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- CNPeking
Shanghang Zhang
- Vision-language & multimodal
- DETUM
Rüdiger Westermann
- Rendering & appearance
- CHEPFL
Pascal Fua
- 3D vision & neural rendering
- CNTsinghua
Hang Zhao
- Embodied & active vision
- CNSJTU
Songhua Liu
- Generative & foundation models
- CNPeking
Yadong Mu
- Video understanding
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
SJTUcloses